Scraping therapy parameter self-adaptive optimization method and system combined with multi-source data

Abstract

The invention discloses a scraping therapy parameter adaptive optimization method and system combined with multi-source data, and relates to the technical field of parameter optimization, and the method comprises the steps: generating a preliminary scraping therapy parameter scheme; a staged scraping therapy target effect is set for each scraping therapy stage; executing the scraping operation in the current scraping stage, and collecting operation dynamic physiological feedback data and user subjective feedback data; evaluating to obtain an effect achievement degree score of the current stage; based on the effect achievement degree score, multi-target self-adaptive optimization is conducted on the scraping therapy parameters of the next scraping therapy stage, the optimal scraping therapy parameters of the next scraping therapy stage are generated, and the scraping therapy instruction is executed. The technical problems that the subjectivity is high and the effect is unstable due to the fact that existing scraping therapy parameters depend on artificial experience judgment, and the personalized adaptability and the intelligent degree are low due to the lack of a dynamic monitoring and real-time feedback mechanism are solved.

Description

Technical Field

[0001] This application relates to the field of parameter optimization technology, specifically to an adaptive optimization method and system for scraping parameters that combines multi-source data. Background Technology

[0002] With the increasing awareness of health, traditional Chinese medicine physiotherapy methods such as scraping are widely used due to their significant effects in relieving fatigue and unblocking meridians.

[0003] However, current gua sha practices largely rely on the operator's experience and judgment, lacking objective and precise parameter guidance. Due to individual differences in physical condition, variations in the physiological structure of the scraping area, and the operator's strong subjectivity in controlling parameters such as force, angle, and speed, it is easy to lead to poor gua sha results or adverse reactions.

[0004] Furthermore, existing technologies lack dynamic monitoring and real-time feedback mechanisms for the scraping process, making it impossible to adjust scraping parameters in a timely manner based on changes in the user's physiological state and subjective feelings during the scraping process. This results in a low level of intelligence and personalization of the entire scraping process, making it difficult to meet the precise health needs of different users. Summary of the Invention

[0005] This application provides an adaptive optimization method and system for scraping parameters that combines multi-source data. This solves the technical problems of existing scraping parameters relying on human experience for judgment, resulting in strong subjectivity and unstable effects, as well as the lack of dynamic monitoring and real-time feedback mechanisms, which leads to low personalization adaptability and intelligence.

[0006] The technical solution to the above-mentioned technical problems in this application is as follows: Firstly, this application provides an adaptive optimization method for scraping parameters combining multi-source data, the method comprising: Collect individual characteristic information of the target user and initial physiological characteristic information of the target scraping area to generate a preliminary scraping parameter scheme; The preset scraping process is divided into multiple scraping stages, and a stage-specific scraping target effect is set for each scraping stage; Based on the preliminary scraping parameter scheme, the scraping operation of the current scraping stage is performed, and dynamic physiological feedback data and user subjective feedback data during the operation process are collected in real time through multi-source sensors. The dynamic physiological feedback data and user subjective feedback data, combined with the actual scraping parameters of the current scraping stage, are input into the scraping effect prediction model for evaluation, and compared with the scraping target effect to obtain the effect achievement score of the current stage. Based on the achievement score, the scraping parameters for the next scraping stage are optimized using multi-objective adaptive optimization to generate the optimal scraping parameters for the next scraping stage, and the scraping command is executed based on the optimal scraping parameters.

[0007] Secondly, this application provides an adaptive optimization system for scraping parameters that combines multi-source data, including: The information acquisition module is used to collect individual characteristic information of the target user and the initial physiological characteristic information of the target scraping area, and generate a preliminary scraping parameter scheme; The stage division module is used to divide the preset scraping process into multiple scraping stages and set the stage-specific scraping target effect for each scraping stage. The scraping execution module is used to execute the scraping operation of the current scraping stage based on the preliminary scraping parameter scheme, and at the same time collect dynamic physiological feedback data and user subjective feedback data in real time through multi-source sensors. The model training module is used to input the dynamic physiological feedback data and user subjective feedback data, combined with the actual scraping parameters of the current scraping stage, into the scraping effect prediction model for evaluation, and compare it with the scraping target effect to obtain the effect achievement score of the current stage. The parameter optimization module is used to perform multi-objective adaptive optimization of the scraping parameters for the next scraping stage based on the effect achievement score, generate the optimal scraping parameters for the next scraping stage, and execute the scraping command based on the optimal scraping parameters.

[0008] This application provides one or more technical solutions, which have at least the following technical effects or advantages: This application provides a method and system for adaptive optimization of scraping parameters by combining multi-source data. First, it collects individual characteristic information of the target user and initial physiological characteristic information of the target scraping area. Second, it divides the entire scraping process into coherent stages and sets quantifiable target effects for each stage, making the scraping process controllable and orderly. Third, it captures dynamic physiological feedback data in real time through multi-source sensors, and combines this with subjective feedback data such as pain scores and comfort scores input by the user through an interactive interface to output a quantified score. Finally, based on this effect achievement score, a multi-objective adaptive optimization algorithm is activated to generate and evaluate the comprehensive fitness of multiple candidate scraping parameter schemes, iteratively optimizing the process, and finally outputting the parameter scheme with the highest comprehensive fitness as the optimal scraping parameter for the next scraping stage. The parameters are dynamically adjusted according to the user's real-time state during the scraping process, realizing intelligent operation of the entire scraping parameter process from initial setting to dynamic optimization.

[0009] Through the above technical solution, this application effectively avoids the subjectivity and limitations of human experience, significantly improves the stability of the scraping effect, the ability to personalize and adapt, and the overall level of intelligence, thereby better meeting the precise health needs of different users. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a flowchart illustrating the adaptive optimization method for scraping parameters that combines multi-source data, provided in an embodiment of this application. Figure 2 This is a schematic diagram of the structure of the adaptive optimization system for scraping parameters that combines multi-source data, provided in an embodiment of this application.

[0012] The components represented by each number in the attached diagram are explained below: Information acquisition module 11, stage division module 12, scraping execution module 13, model training module 14, parameter optimization module 15. Detailed Implementation

[0013] This application provides a method and system for adaptive optimization of scraping parameters by combining multi-source data. This method addresses the technical problems of existing scraping parameters relying on human experience, which leads to strong subjectivity and unstable effects, as well as the lack of dynamic monitoring and real-time feedback mechanisms, resulting in low personalization adaptability and intelligence.

[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0015] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0016] In the description of this application, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid unnecessarily obscuring the description of this application. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0017] Example 1, as Figure 1 As shown in the embodiments of this application, an adaptive optimization method for scraping parameters combining multi-source data is provided, including: S10: Collect individual characteristic information of the target user and initial physiological characteristic information of the target scraping area to generate a preliminary scraping parameter scheme; The initial scraping parameter scheme includes the initial scraping force, initial scraping angle, initial scraping speed, and initial single scraping duration.

[0018] In this embodiment of the application, firstly, the individual characteristic information of the target user is collected in multiple dimensions, including age, gender, height, weight, body type, history of scraping therapy, current health status, and daily exercise frequency and intensity.

[0019] Meanwhile, for the target scraping area, sensors collect its initial physiological characteristics, including the base skin temperature, initial skin impedance value, initial subcutaneous soft tissue hardness, local initial blood flow velocity, and skin color baseline data.

[0020] Furthermore, the collected individual characteristic information is integrated with the initial physiological characteristic information, and combined with the scraping characteristics of different parts in the theory of meridians in traditional Chinese medicine, a preliminary scraping parameter scheme is generated, which includes the initial scraping force, initial scraping angle, initial scraping speed, and initial single scraping duration.

[0021] S20: Divide the preset scraping process into multiple scraping stages, and set a stage-specific scraping target effect for each scraping stage; The scraping process includes a skin adaptation and preheating stage, a core petechiae emergence stage, and a soothing and finishing stage.

[0022] In this embodiment, the pre-set scraping process is divided into three coherent and progressive stages according to the physiological reaction law of the human body and the scraping treatment logic: the skin adaptation and preheating stage, the core sha emergence stage, and the soothing and finishing stage.

[0023] Secondly, for different stages of scraping, set phased scraping target effects and quantify the indicator system.

[0024] Furthermore, set phased gua sha target effects for each stage, including: The target effect of the skin adaptation and preheating stage is when the local skin temperature rises to 1.05-1.1 times the basal body temperature and the skin color reaches the first preset color. The scraping area shows evenly distributed light red to purplish-red petechiae, and the local blood flow velocity increases by 30%-50% compared to the initial value, which is the target effect of the core petechiae stage. The skin surface petechiae no longer darken and the skin temperature drops by 5%-8% from the peak of the core petechiae stage, which is the target effect of the soothing and closing stage.

[0025] In this embodiment, the target effect of the skin adaptation and preheating stage is that, at the end of this stage, the local skin temperature of the target scraping area is monitored by an infrared temperature sensor to rise by 1.05-1.1 times compared to the basal body temperature. For example, if the user's basal body temperature is 36.5℃, the target temperature range is 38.325℃-39.15℃. At the same time, the skin color change is analyzed by an image acquisition device to achieve a first preset color, which is usually set as a slight flush. Specifically, it can be defined by a specific threshold range in the RGB color model, such as an R value between 230-245, a G value between 180-200, and a B value between 170-190, to ensure that the skin and subcutaneous tissue are fully preheated, preparing for the subsequent core petechiae stage, while avoiding user discomfort due to excessive initial stimulation.

[0026] Secondly, the target effect of the core stage of petechiae treatment is set as the appearance of evenly distributed petechiae on the scraping area, with the color of the petechiae transitioning from light red to purplish red. Furthermore, based on existing image recognition technology, the area of ​​the petechiae needs to cover 40%-60% of the target scraping area to avoid excessive density or sparseness in certain areas. At the same time, through laser Doppler blood flow monitoring, the local blood flow velocity in this area is increased by 30%-50% compared to the initial value collected in the initial physiological characteristic information, thereby quantitatively evaluating the effect of meridian unblocking and improvement of local blood circulation.

[0027] Finally, the goal of the soothing and finishing stages is to achieve a smooth transition and reduce discomfort. Specifically, the color of the petechiae already formed on the skin surface should no longer deepen, and the skin temperature should decrease by 5%-8% from the peak temperature during the core petechiae-forming stage, as monitored by an infrared temperature sensor. For example, if the peak temperature during the core stage is 39℃, the target temperature during the finishing stage should be between 36.8℃ and 37.1℃. Simultaneously, the pain score reported by the user through the interactive interface should decrease to 2 points or below. Here, 0 points represents no pain, and 10 points represents the most severe pain. This ensures that while achieving the therapeutic effect, the user's comfort is maximized, promoting initial recovery of the scraped area.

[0028] S30: Based on the preliminary scraping parameter scheme, perform the scraping operation of the current scraping stage, and at the same time collect dynamic physiological feedback data and user subjective feedback data in real time through multi-source sensors. The dynamic physiological feedback data includes data on changes in skin impedance, changes in local skin temperature field, and changes in subcutaneous soft tissue stiffness. User subjective feedback data includes user pain ratings and comfort ratings.

[0029] In this embodiment of the application, when performing the scraping operation in the current scraping stage, the scraping execution device is controlled to act on the target scraping area according to the initial scraping force, angle, speed and duration parameters based on the preliminary scraping parameter scheme generated in S10.

[0030] Meanwhile, in order to achieve dynamic monitoring of the scraping process, a multi-source sensor system is deployed for real-time data acquisition.

[0031] Specifically, the system continuously monitors changes in skin impedance at the target scraping site using a skin impedance sensor. This data reflects the dynamic changes in skin sweat gland activity and local blood circulation. An infrared thermal imaging sensor collects local skin temperature field change data to generate a temperature field distribution map, which tracks the spatial distribution of skin temperature and its evolution over time in real time. A pressure sensor array or an ultrasonic elastography sensor is used to obtain data on changes in subcutaneous soft tissue hardness, which quantitatively assesses changes in the tension and relaxation of subcutaneous tissue during the scraping process.

[0032] In addition, a user interface is set up so that target users can input subjective feedback data in real time through a touch screen, voice input or a dedicated handheld controller. This includes a pain score of 0-10 and a comfort score of 0-5. Users can adjust and submit their feedback at any time during the scraping process based on their own feelings, thus forming a multi-source feedback information that combines objective physiological data and subjective feeling data.

[0033] S40: The dynamic physiological feedback data and user subjective feedback data, combined with the actual scraping parameters of the current scraping stage, are input into the scraping effect prediction model for evaluation, and compared with the scraping target effect to obtain the effect achievement score of the current stage. In this embodiment of the application, firstly, a scraping effect prediction model is constructed, which adopts a fusion architecture of multilayer perceptron and long short-term memory network.

[0034] The LSTM layer is used to process the temporal features in the dynamic physiological feedback data, such as the trend of skin impedance changes, the evolution sequence of the temperature field over time, and the dynamic fluctuation process of subcutaneous soft tissue hardness; the MLP layer is used to integrate the user's subjective feedback data with the actual scraping parameters at the current stage.

[0035] Furthermore, during the evaluation process, the dynamic physiological feedback data collected by S30 is preprocessed, including data cleaning, normalization, and time-series feature extraction; at the same time, the subjective feedback data of users is standardized, and the pain numerical score and comfort score are mapped to the [0,1] interval.

[0036] The model outputs a predicted value of the scraping effect at the current stage by fusing features and nonlinear mapping the input data. This predicted value includes a quantitative indicator corresponding to the target effect at the current scraping stage. Subsequently, the predicted value is compared with the preset target effect of scraping one by one, and the effect achievement score is calculated by weighted summation.

[0037] The scraping effect prediction model is constructed through the following steps: Based on the historical gua sha case database, the sample user individual characteristics, sample gua sha site characteristics, sample gua sha parameters, sample gua sha effect evaluation, and sample adverse reaction records were collected. The evaluation of the effects of scraping on the samples included the degree of petechiae and the degree of improvement in local microcirculation, which served as the sample scraping effect index set; the adverse reaction records of the samples included the probability of skin damage and the peak pain level, which served as the sample risk index set. Using sample scraping parameters, individual characteristics of sample users, and characteristics of sample scraping sites as inputs, and using sample scraping effect index set and sample risk index set as supervision, a neural network model is trained to obtain the scraping effect prediction model.

[0038] In this embodiment of the application, firstly, sample data that meets clinical standards and is complete is selected from the historical Gua Sha case database. Each sample includes the individual characteristics of the sample user, the characteristics of the sample scraping site, the sample scraping parameters, the effect evaluation after the sample scraping, and the sample adverse reaction record.

[0039] The evaluation of the effects of scraping after the sample was conducted by professional physicians using a standard scale, which included the degree of petechiae and the degree of improvement in local microcirculation, forming the sample scraping effect index set; the adverse reaction records included the probability of skin damage such as breakage and bruising, as well as the peak pain reported by users, forming the sample risk index set.

[0040] Secondly, the sample scraping parameters, individual characteristics of sample users, and characteristics of sample scraping sites are used as input variables of the model. The sample scraping effect index set and sample risk index set are used as supervision signals, and the neural network model is trained by multi-task learning.

[0041] During model training, the network weights are continuously adjusted and the loss function is optimized using the backpropagation algorithm. This loss function comprehensively considers the prediction error of the effect index and the penalty term of the risk index, aiming to improve the effect of Gua Sha while minimizing the probability of adverse reactions. After multiple rounds of iterative training and validation, training stops when the model's prediction accuracy and stability reach a preset threshold, resulting in the final Gua Sha effect prediction model. This model can accurately predict the effect and potential risks of Gua Sha based on the input parameters and features.

[0042] For example, the steps for building and training a gua sha effect prediction model based on a neural network are as follows: First, data preparation involves screening sample scraping parameters, individual characteristics of sample users and characteristics of the scraping sites, sample scraping parameters, post-scraping effect evaluation, and records of adverse reactions based on a historical scraping case database.

[0043] Secondly, the model was constructed using an LSTM-MLP fusion network. The LSTM part contains two hidden layers: the first layer has 64 neurons and the second layer has 32 neurons, using the ReLU activation function to handle potential temporal correlations in the initial physiological features, and a Dropout layer is used to prevent overfitting. The MLP part contains three fully connected layers with 128, 64, and 32 neurons respectively, also using the ReLU activation function to integrate individual user features with static features of the scraping site. The model output layer has five neurons, corresponding to the degree of petechiae (0-10 points), the degree of microcirculation improvement (0-10 points), the probability of skin damage (0-1), the pain peak (0-10 points), and the effect achievement score (0-100 points).

[0044] Next, for model training, the Adam optimizer was used with an initial learning rate of 0.001, which decayed by 10% every 10 epochs, for a total of 100 training epochs. The loss function was a weighted sum of mean squared error (MSE) and cross-entropy loss, with MSE loss used as the performance metric and cross-entropy loss used as the risk metric, with weights of 0.6 and 0.4, respectively.

[0045] Furthermore, during training, the training set and validation set are divided in an 8:2 ratio, and the validation set loss is monitored in real time. When the validation set loss does not decrease for 15 consecutive epochs, an early stopping mechanism is triggered to save the current optimal model parameters.

[0046] Finally, the model was evaluated using mean absolute error, root mean square error, and accuracy as evaluation metrics. The goal was to ensure that the mean absolute error of the model on the test set did not exceed 0.8 points, the accuracy of skin lesion probability prediction reached over 90%, and the root mean square error of the overall performance score was controlled within 5 points, thus guaranteeing the reliability of the model's predictions.

[0047] Furthermore, the achievement score is calculated by comparing each quantitative indicator in the predicted value with the corresponding indicator of the target effect in the current stage, and assigning different weights to each indicator according to its importance in this stage. The actual achievement ratio of each indicator is multiplied by its corresponding weight and then summed to obtain the actual effect score of the current stage. The score range is 0-100 points, with 60 points being basic achievement, 80 points being good achievement, and 90 points and above being excellent achievement.

[0048] For example, during the skin adaptation and preheating phase, if the skin temperature in the current dynamic physiological feedback data reaches 38.325℃ from the basal body temperature of 36.5℃, which is 1.05 times the basal body temperature, and the skin color R value is within the range of 230-245, G value 180-200, and B value 170-190, while the user's pain score is 1 and comfort score is 4, then the percentage of skin temperature increase achieved is (current temperature - basal body temperature) / (target temperature upper limit - basal body temperature). If the current temperature is 38.5℃, then (38.5-36.5) / (39.15-36.5) ≈ 0.754, and the percentage of skin color achievement is 1. Since it is already within the target range, the effect achievement score = 0.754 × 0.6 + 1 × 0.4 ≈ 0.852, which is 85.2 points.

[0049] S50: Based on the effect achievement score, perform multi-objective adaptive optimization of the scraping parameters for the next scraping stage, generate the optimal scraping parameters for the next scraping stage, and execute the scraping command based on the optimal scraping parameters.

[0050] In this embodiment of the application, firstly, a multi-objective optimization function is set, with the core optimization objectives being to improve the degree of effect achievement, reduce user pain scores, and maintain operational safety.

[0051] Secondly, parameter optimization constraints are constructed, and parameter adjustment ranges are set according to the characteristics of different scraping stages. The effect of each candidate parameter combination is pre-evaluated by the scraping effect prediction model. The predicted effect achievement, pain score and damage risk are substituted into the multi-objective optimization function to calculate the fitness value. The parameter combination with the highest fitness value is selected as the optimal scraping parameters for the next stage.

[0052] Furthermore, the generated optimal scraping parameters are converted into specific execution instructions and sent to the drive module of the scraping execution device. The motor is controlled to adjust the speed of the eccentric wheel to change the scraping force, the servo motor is used to adjust the tilt angle of the scraping head, the conveyor belt or robotic arm is driven to move to set the scraping speed, and the timer is started to control the duration of a single scraping.

[0053] Specifically, step S50 in the method includes: Within a preset parameter perturbation range, a first candidate scraping parameter scheme is generated, and the comprehensive fitness of the candidate scraping parameter scheme is calculated as the first comprehensive fitness. Within a preset parameter perturbation range, a second candidate scraping parameter scheme is generated, and the comprehensive fitness of the candidate scraping parameter scheme is calculated as the second comprehensive fitness. Iterate through the candidate scraping parameter schemes, perform iterative optimization, and output the scraping parameter scheme with the highest comprehensive adaptability as the optimal scraping parameter for the next scraping stage; Update the control instructions of the scraping device according to the optimal scraping parameters, perform the scraping operation of the next scraping stage, and iterate and optimize the parameters according to the order of the scraping stages until all scraping stages are completed.

[0054] In this embodiment, firstly, within a preset parameter perturbation range, an adjustment range is set for each scraping parameter in the current stage. For example, the perturbation range for scraping force can be set to ±15% of the current parameter value, the perturbation range for scraping angle is ±5°, the perturbation range for scraping speed is ±10%, and the perturbation range for scraping duration is ±20%. Based on this perturbation range, multiple first candidate scraping parameter schemes are generated using methods such as random sampling or Latin hypercube sampling.

[0055] For each first candidate solution, it is input into the pre-trained gua sha effect prediction model, and the model will output the corresponding prediction effect indicators, including the effect achievement score, the predicted pain score, and the probability of skin damage.

[0056] Subsequently, the overall fitness of the candidate solution is calculated according to the preset multi-objective optimization function. This optimization function can be expressed as: Overall fitness = α × effect achievement score + β × (1 - normalized pain score) + γ × (1 - skin damage probability), where α, β, and γ are the weight coefficients of each objective, which can be adjusted according to the focus of the current scraping stage. For example, in the core scraping stage, the weight of α can be appropriately increased, while in the soothing and finishing stage, the weights of β and γ can be increased accordingly to prioritize user comfort and safety.

[0057] Next, using the same method, a second candidate scraping parameter scheme is generated within the preset parameter perturbation range, and its comprehensive fitness is calculated. Here, "second" does not specifically refer to a second in quantity, but rather represents a new round of candidate scheme generation to increase the diversity and coverage of the parameter search. In practice, the number of candidate schemes generated can be set according to computational resources and optimization efficiency requirements, for example, generating 5-10 sets of candidate schemes each time.

[0058] Then, all generated candidate scraping parameter schemes are traversed, including the first, second, and subsequent candidate schemes, and their overall fitness values ​​are compared. Through iterative optimization, such as selection, crossover, and mutation operations in genetic algorithms, the candidate parameter space is continuously evolved, eliminating schemes with lower fitness and retaining and optimizing those with higher fitness. During this process, the parameter perturbation range is dynamically adjusted, gradually narrowing as the number of iterations increases, making the search more focused on the potential optimal solution region. After a preset number of iterations, or when the improvement in the overall fitness value is less than a certain threshold (e.g., 0.5%) after multiple consecutive iterations, iteration stops, and the scraping parameter scheme with the highest overall fitness is determined as the optimal scraping parameter for the next scraping stage.

[0059] Finally, the determined optimal scraping parameters are parsed into specific control commands for each drive component of the scraping execution device. For example, the scraping force corresponds to the output torque of the motor or the pressure value of the pneumatic cylinder, the scraping angle corresponds to the rotation angle of the servo motor, the scraping speed corresponds to the linear speed of the conveyor belt or the movement speed of the robotic arm joint, and the scraping duration corresponds to the set value of the timer. The control commands are sent to the main controller of the scraping device through the communication module. The main controller drives the corresponding actuators according to the commands to perform the scraping operation of the next stage.

[0060] In summary, compared with existing technologies, this application constructs a scraping effect prediction model by combining multi-source data, and realizes multi-objective adaptive optimization of scraping parameters based on the model, thereby achieving intelligent and personalized scraping process.

[0061] In summary, the embodiments of this application have at least the following technical effects: This application provides an adaptive optimization method for scraping parameters that combines multi-source data. First, it collects individual characteristic information of the target user and initial physiological characteristic information of the target scraping area. Second, it divides the entire scraping process into coherent stages and sets quantifiable target effects for each stage, making the scraping process controllable and orderly. Third, it captures dynamic physiological feedback data in real time through multi-source sensors, and combines this with subjective feedback data such as pain scores and comfort scores input by the user through an interactive interface to output a quantified score. Finally, based on this effect achievement score, a multi-objective adaptive optimization algorithm is activated to generate and evaluate the comprehensive fitness of multiple candidate scraping parameter schemes, iteratively optimizing the process, and finally outputting the parameter scheme with the highest comprehensive fitness as the optimal scraping parameter for the next scraping stage. The parameters are dynamically adjusted according to the user's real-time state during the scraping process, realizing intelligent operation of the entire scraping parameter process from initial setting to dynamic optimization.

[0062] Through the above technical solution, this application effectively avoids the subjectivity and limitations of human experience, significantly improves the stability of the scraping effect, the ability to personalize and adapt, and the overall level of intelligence, thereby better meeting the precise health needs of different users.

[0063] Example 2, as Figure 2 As shown, based on the same inventive concept as the scraping parameter adaptive optimization method combining multi-source data provided in Embodiment 1, this application also provides a scraping parameter adaptive optimization system combining multi-source data, including: The information acquisition module 11 is used to collect the individual characteristic information of the target user and the initial physiological characteristic information of the target scraping area, and generate a preliminary scraping parameter scheme. The stage division module 12 is used to divide the preset scraping process into multiple scraping stages and set stage-specific scraping target effects for each scraping stage. The scraping execution module 13 is used to execute the scraping operation of the current scraping stage based on the preliminary scraping parameter scheme, and at the same time collect dynamic physiological feedback data and user subjective feedback data in real time through multi-source sensors. The model training module 14 is used to input the dynamic physiological feedback data and user subjective feedback data, combined with the actual scraping parameters of the current scraping stage, into the scraping effect prediction model for evaluation, and compare it with the scraping target effect to obtain the effect achievement score of the current stage. The parameter optimization module 15 is used to perform multi-objective adaptive optimization of the scraping parameters for the next scraping stage based on the effect achievement score, generate the optimal scraping parameters for the next scraping stage, and execute the scraping command based on the optimal scraping parameters.

[0064] Furthermore, in one embodiment of the application, the initial scraping parameter scheme includes initial scraping force, initial scraping angle, initial scraping speed, and initial single scraping duration.

[0065] Furthermore, in one embodiment of the application, the scraping stage includes a skin adaptation and preheating stage, a core petechiae emergence stage, and a soothing and finishing stage.

[0066] Among them, setting phased gua sha target effects for each stage includes: The target effect of the skin adaptation and preheating stage is when the local skin temperature rises to 1.05-1.1 times the basal body temperature and the skin color reaches the first preset color. The scraping area shows evenly distributed light red to purplish-red petechiae, and the local blood flow velocity increases by 30%-50% compared to the initial value, which is the target effect of the core petechiae stage. The skin surface petechiae no longer darken and the skin temperature drops by 5%-8% from the peak of the core petechiae stage, which is the target effect of the soothing and closing stage.

[0067] In one embodiment, dynamic physiological feedback data includes skin impedance change data, local skin temperature field change data, and subcutaneous soft tissue stiffness change data. User subjective feedback data includes user pain ratings and comfort ratings.

[0068] Furthermore, in one embodiment, the scraping effect prediction model is constructed through the following steps: Based on the historical gua sha case database, the sample user individual characteristics, sample gua sha site characteristics, sample gua sha parameters, sample gua sha effect evaluation, and sample adverse reaction records were collected. The evaluation of the effects of scraping on the samples included the degree of petechiae and the degree of improvement in local microcirculation, which served as the sample scraping effect index set; the adverse reaction records of the samples included the probability of skin damage and the peak pain level, which served as the sample risk index set. Using sample scraping parameters, individual characteristics of sample users, and characteristics of sample scraping sites as inputs, and using sample scraping effect index set and sample risk index set as supervision, a neural network model is trained to obtain the scraping effect prediction model.

[0069] In one embodiment, the parameter optimization module 15 is specifically used for: Within a preset parameter perturbation range, a first candidate scraping parameter scheme is generated, and the comprehensive fitness of the candidate scraping parameter scheme is calculated as the first comprehensive fitness. Within a preset parameter perturbation range, a second candidate scraping parameter scheme is generated, and the comprehensive fitness of the candidate scraping parameter scheme is calculated as the second comprehensive fitness. Iterate through the candidate scraping parameter schemes, perform iterative optimization, and output the scraping parameter scheme with the highest comprehensive adaptability as the optimal scraping parameter for the next scraping stage; Update the control instructions of the scraping device according to the optimal scraping parameters, perform the scraping operation of the next scraping stage, and iterate and optimize the parameters according to the order of the scraping stages until all scraping stages are completed.

[0070] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0071] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0072] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. An adaptive optimization method for scraping parameters based on multi-source data, characterized in that, The method includes: Collect individual characteristic information of the target user and initial physiological characteristic information of the target scraping area to generate a preliminary scraping parameter scheme; The preset scraping process is divided into multiple scraping stages, and a stage-specific scraping target effect is set for each scraping stage; Based on the preliminary scraping parameter scheme, the scraping operation of the current scraping stage is performed, and dynamic physiological feedback data and user subjective feedback data during the operation process are collected in real time through multi-source sensors. The dynamic physiological feedback data and user subjective feedback data, combined with the actual scraping parameters of the current scraping stage, are input into the scraping effect prediction model for evaluation, and compared with the scraping target effect to obtain the effect achievement score of the current stage. Based on the achievement score, the scraping parameters for the next scraping stage are optimized using multi-objective adaptive optimization to generate the optimal scraping parameters for the next scraping stage, and the scraping command is executed based on the optimal scraping parameters.

2. The adaptive optimization method for scraping parameters combining multi-source data according to claim 1, characterized in that, The initial scraping parameter scheme includes initial scraping force, initial scraping angle, initial scraping speed, and initial single scraping duration.

3. The adaptive optimization method for scraping parameters combining multi-source data according to claim 1, characterized in that, The scraping process includes the skin adaptation and preheating stage, the core petechiae emergence stage, and the soothing and finishing stage.

4. The adaptive optimization method for scraping parameters combining multi-source data according to claim 3, characterized in that, Set phased gua sha target effects for each stage, including: The target effect of the skin adaptation and preheating stage is when the local skin temperature rises to 1.05-1.1 times the basal body temperature and the skin color reaches the first preset color. The scraping area shows evenly distributed light red to purplish-red petechiae, and the local blood flow velocity increases by 30%-50% compared to the initial value, which is the target effect of the core petechiae stage. The skin surface petechiae no longer darken and the skin temperature drops by 5%-8% from the peak of the core petechiae stage, which is the target effect of the soothing and closing stage.

5. The adaptive optimization method for scraping parameters combining multi-source data according to claim 1, characterized in that, in, Dynamic physiological feedback data includes data on changes in skin impedance, changes in local skin temperature field, and changes in subcutaneous soft tissue stiffness; User subjective feedback data includes user pain ratings and comfort ratings.

6. The adaptive optimization method for scraping parameters combining multi-source data according to claim 1, characterized in that, The scraping effect prediction model is constructed through the following steps: Based on the historical gua sha case database, the sample user individual characteristics, sample gua sha site characteristics, sample gua sha parameters, sample gua sha effect evaluation, and sample adverse reaction records were collected. The evaluation of the effects of scraping on the samples included the degree of petechiae and the degree of improvement in local microcirculation, which served as the sample scraping effect index set; the adverse reaction records of the samples included the probability of skin damage and the peak pain level, which served as the sample risk index set. Using sample scraping parameters, individual characteristics of sample users, and characteristics of sample scraping sites as inputs, and using sample scraping effect index set and sample risk index set as supervision, a neural network model is trained to obtain the scraping effect prediction model.

7. The adaptive optimization method for scraping parameters combining multi-source data according to claim 1, characterized in that, Based on the achievement score, the scraping parameters for the next scraping stage are optimized using multi-objective adaptive methods to generate the optimal scraping parameters for the next scraping stage. Scraping instructions are then executed based on these optimal parameters, including: Within a preset parameter perturbation range, a first candidate scraping parameter scheme is generated, and the comprehensive fitness of the candidate scraping parameter scheme is calculated as the first comprehensive fitness. Within a preset parameter perturbation range, a second candidate scraping parameter scheme is generated, and the comprehensive fitness of the candidate scraping parameter scheme is calculated as the second comprehensive fitness. Iterate through the candidate scraping parameter schemes, perform iterative optimization, and output the scraping parameter scheme with the highest comprehensive adaptability as the optimal scraping parameter for the next scraping stage; Update the control instructions of the scraping device according to the optimal scraping parameters, perform the scraping operation of the next scraping stage, and iterate and optimize the parameters according to the order of the scraping stages until all scraping stages are completed.

8. A scraping sha parameter adaptive optimization system combining multi-source data, characterized in that, The method for adaptive optimization of scraping parameters combining multi-source data as described in any one of claims 1-7 includes: The information acquisition module is used to collect individual characteristic information of the target user and the initial physiological characteristic information of the target scraping area, and generate a preliminary scraping parameter scheme; The stage division module is used to divide the preset scraping process into multiple scraping stages and set the stage-specific scraping target effect for each scraping stage. The scraping execution module is used to execute the scraping operation of the current scraping stage based on the preliminary scraping parameter scheme, and at the same time collect dynamic physiological feedback data and user subjective feedback data in real time through multi-source sensors. The model training module is used to input the dynamic physiological feedback data and user subjective feedback data, combined with the actual scraping parameters of the current scraping stage, into the scraping effect prediction model for evaluation, and compare it with the scraping target effect to obtain the effect achievement score of the current stage. The parameter optimization module is used to perform multi-objective adaptive optimization of the scraping parameters for the next scraping stage based on the effect achievement score, generate the optimal scraping parameters for the next scraping stage, and execute the scraping command based on the optimal scraping parameters.

Images